We are coming down to the last week for JTC1 to decide on whether to hear the four NB appeals concerning various claimed errors in the processing of DIS 29500 (OOXML), or whether summarily to dismiss these appeals without hearing them. The decision lies with two committees, the Technical Management Board (TMB) in ISO and the Standards Management Board (SMB) in the IEC.
Back on July 4th, the Secretaries General of ISO and the IEC referred the four NB appeals, with their comments, to the TMB/SMB. Groklaw has the text of these comments, in PDF format, as well as HTML transcription.
The comments of the Secretary General are accompanied by a ballot, asking the question:
The members of the Technical Management Board are invited to indicate, by replying yes, no or
abstention on EITHER a) OR b) for each of the four appeals (see item 14 in annex A):
a) not to process the appeal any further:
Item 1 ABNT
Item 2 BIS
Item 3 FONDONORMA
Item 4 SABS
b) to process one or more of the appeals, which would require setting up of a conciliation panel
Item 5 ABNT
Item 6 BIS
Item 7 FONDONORMA
Item 8 SABS
by no later than 4 August 2008.
This is quite a strange animal to see. Why are we having a ballot at all, and only a 30-day one? This is questionable from several perspectives.
First, why are the Secretaries General the ones calling for a ballot? The Directives do not call for them to do so. In fact the Secretaries General are not even called upon to make a recommendation. They are only asked for comments. The Directives say:
11.3.3 The Secretaries-General shall, following whatever consultations they deem appropriate, refer the appeal together with their comments to the TMB/SMB within one month after receipt of the appeal.
11.3.4 The TMB/SMB shall decide whether an appeal shall be further processed or not. If the decision is in favour of proceeding, the Chairmen of the TMB/SMB shall form a conciliation panel (see 9.2).
But deciding is not the same as voting. One of the cardinal principles of JTC1 is to discuss and seek consensus, not rush to a vote. Indeed, this is one of the matters under appeal, the rush to voting at the OOXML BRM. JTC1 Directives, section 1.2 says (my emphasis):
These Directives are inspired by the principle that the objective in the development of International Standards should be the achievement of consensus between those concerned rather than a decision based on counting votes.
But here we are, with a vote pushed on the TMB/SMB.
The sense of the vote is wrong as well. The Directives call for a decision on “whether an appeal shall be further processed or not.” Note the wording. It did not call for a decision on “whether to accept the recommendation of the Secretaries General”. But somehow, we skip discussion, skip over consensus and get a ballot question which asks the opposite question first “not to process the appeal any further”. In an environment where many parties automatically vote Yes to the ballot question, changing the sense of the question in this way is prejudicial to the appellants.
So it is clear from the start that the powers that be do not want to give these four NB’s the opportunity to make their case or be heard. In any case, let’s take a deeper look at some of the subjects under appeal and see if we can detect what it is exactly that cannot bear the scrutiny of a duly processed appeal.
First up is the alleged mishandling of the contradiction period last year. The Secretaries General dismiss this complaint, saying that it was a matter of judgment:
The Directives give the JTC 1 Secretariat and ITTF latitude to use judgement as to whether a meeting should be organized to address alleged contradictions. Considering that other issues could potentially be identified during the DIS ballot, the JTC 1 secretariat and ITTF concluded that it was preferable to initiate the ballot and to allow all issues to be addressed by the BRM. The NBs were fully informed of all the claimed contradictions and Ecma’s responses to them.
This argument doesn’t hold water. Although the JTC1 Secretariat and ITTF are allowed judgment, this is not an absolute license which cannot be questioned. The Secretariat and ITTF also have defined duties, and their actions or inactions with respect to these duties can be questioned and are subject to appeal. Specifically, an NB may appeal the issue of an inaction of JTC1, according to JTC1 Directives, 11.3. So for the Secretaries General to suggest that this inaction cannot be appealed because it is a matter of judgment is nonsense. Judgment and duty are the proper matters for an appeal.
So what is the duty in this case? As stated in JTC1 Directives, 13.4:
If a contradiction is alleged, the JTC 1 Secretariat and ITTF shall make a best effort to resolve the matter in no more than a three month period, consulting with the proposer of the fast-track document, the NB(s) raising the claim of contradiction and others, as they deem necessary. A meeting of these parties, open to all NBs, may be convened by the JTC 1 Secretariat, if required.
If the resolution requires a change to the document submitted for fast-track processing, the initial document submitted will be considered withdrawn. The proposer may submit a revised document, to be processed as a new proposal.
If the resolution results in no change to the document or if a resolution cannot be reached, the five month fast-track ballot commences immediately after such a determination is made.
The Directives call for the JTC1 Secretariat to make a best effort to resolve the matter (JTC1 Directives, 13.4). The JTC1 Secretariat is not given latitude to do nothing, or allowed discretion to immediately defer this question to the ballot period, without making a best effort to resolve the matter.
When a new 6,000 page DIS is submitted to JTC1 only one month after the publication of another standard (ODF) in the exact same space (XML document formats for office applications) and 19 NB’s submit contradiction statements, and the JTC1 Secretariat’s “best effort” is to hold no consultations with the NB’s claiming contradictions, to hold no meeting, to make no attempt to resolve the question, then I believe that any NB would has a legitimate grounds for appeal on the inaction of JTC1 with regards to contradictions. There is no evidence that a “best effort” was made here to resolve the contradictions. Doing nothing is clearly incompatible with the required “best effort”.
It should be noted that JTC1 has had challenges in the past getting ITTF to carry out their responsibilities with respect to contradictions, which lead to this resolution adopted unamimously at the 2000 JTC1 Plenary:
Resolution 27 – Consistency of JTC 1 Products
JTC 1 stresses the strong need for consistency of its products (ISs and TRs) irrespective of the route through which they were developed. Any inconsistency will confuse users of JTC 1 standards and, hence, jeopardize JTC 1’s reputation. Therefore, referring to clauses 13.2 (Fast Track) and 18.104.22.168 (PAS) of its Directives, JTC 1 reminds ITTF of its obligation to ascertain that a proposed DIS contains no evident contradiction with other ISO/IEC standards. JTC 1 offers any help to ITTF in such undertaking. However, should an inconsistency be detected at any point in the ratification process, JTC 1 together with ITTF will take immediate action to cure the problem.
Perhaps it is time to give ITTF another reminder of their obligations in this regard?
Further, the determination claimed to have been made by the JTC1 Secretariat and ITTF was not communicated to JTC1 NB’s. Instead, the JTC1 Secretariat merely forwarded Ecma’s responses to the contradiction submissions along with a notification that the DIS ballot should then commence. No statement was made as to whether the ballot was commencing because the contradictions had in fact been resolved, or because a resolution could not be made, which are the only two outcomes allowed by the Directives in 13.4. Not to notify NB’s of the actual state of the resolution of the contradictions submissions is incompatible with the JTC1 Secretariat’s duty to make a best effort to resolve the matter.
This failure by JTC1 materially effected the ensuing ballot, since Microsoft was then able to take advantage of this procedural nonperformance and repeatedly represent to NB’s that the contradictions had been rejected as invalid and could not be considered in the DIS ballot. In fact, this led to several NB’s issuing explicit, but erroneous instructions to their members that the contradictions had been resolved and thus could not be raised again as a criterion for determining their national position, e.g., in Australia.
Further, although the Secretaries General claim that “the JTC 1 secretariat and ITTF concluded that it was preferable to initiate the ballot and to allow all issues to be addressed by the BRM” the documented fact is that the BRM Convenor explicitly disallowed any discussion of contradictions at the BRM.
Another subject of appeal was the irregular voting procedures used at the DIS 29500 BRM in February. This is the P-member versus O-member question. The Secretaries General dismiss this appeal in this way:
2e. Correct but inapplicable. The BRM was neither a meeting of JTC 1 nor of SC 34 but was open to all 87 national bodies which submitted a vote (including abstentions) on the DIS. Applying 9.1.4 would have disenfranchised the voting NBs present at the BRM which were not P-members. The fact that any votes in the BRM would be open to all national delegations present was communicated over three months prior to the BRM.
This argument presented is flawed, and amounts to saying, “The voting was done by P- and O-members because the meeting was attended by delegations from P- and O-members”. Who attended the meeting is immaterial. Liaisons such as Ecma also attended the BRM? Should they have been able to vote merely because they attended? No, of course not. Voting rights are defined in JTC1 Directives, and this must not be set aside in favor of an ad-hoc rule made without NB consultation or approval.
Asserting that applying 9.1.4 would disenfranchise NB’s is an example of circular reasoning. One can only be disenfranchised if one first has the right to vote. So the statement by the Secretaries General is arguing a conclusion (O-members are permitted to vote at BRM’s) by assuming the very thing it tries to prove.
JTC1 Directives 22.214.171.124, which defines the parallel BRM process for the Publicly Available Specification (PAS) transposition process, reads: “At the ballot resolution group meeting, decisions should be reached preferably by consensus. If a vote is unavoidable, the approval criteria in the subclause 9.1.4 is applied.” So here we see 9.1.4 explicitly called for. By the argument put forth by the Secretaries General, all PAS BRM’s which follow the Directives are also flawed because they “disenfranchise” those NB’s who are not P-members of JTC1. I believe this is a tortured reading of the Directives. The voting rules of 9.1.4 are explicitly and unambiguously called for in PAS BRM’s, so one cannot dismiss their application to Fast Track on general principles that would apply equally to PAS. When Fast Track rules say that the BRM vote shall (“if a vote is unavoidable”) “be taken according to normal JTC 1 procedures” then we are faced with two alternatives:
- Use the voting rules of 9.1.4, which declares itself to be the normal voting procedures (“In a meeting, except as otherwise specified in these directives, questions are decided by a majority of the votes cast at the meeting by P-members expressing either approval or disapproval.”)
- Or use a voting rule which is not to be found anywhere within the Directives.
Which one is “normal JTC 1 procedures”? Where is the basis in the Directives for believing that O-members had the right to vote at the BRM?
Finally, neither BRM Convenor, Alex Brown, nor ITTF, nor indeed the assembled delegations at the BRM were competent nor had the mandate to make or change voting rules for a DIS BRM. The rules are set in JTC1 Directives, and must be followed. “These Directives shall be complied with in all respects and no deviations can be made without the consent of the Secretaries-General.” (JTC1 Directives 1.2).
Notifications made by the BRM Convenor in advance of the BRM have no weight on matters which exceed his mandate and authority. The communication referred to by the Secretaries General, which was given in advance by the BRM Convenor, was from this FAQ:
6.8 If votes are taken during the BRM, who votes?
This in fact was not the rule applied at the BRM. For example, Liaison representatives could not vote, though they were undoubtedly present at the BRM and participated fully in other ways. Also individual participants could not vote, only delegations, via their HoD could vote. So the Convenor’s glib communication should not be taken as notification of a novel voting procedure.
Additionally, the BRM Convenor was unambiguous in his communications on his blog where he clearly stated that the voting rules of 9.1.4 would be applied:
…Now, paper balloting follows normal JTC 1 in-meeting rules: In a meeting, except as otherwise specified in these directives, questions are decided by a majority of the votes cast at the meeting by P-members expressing either approval or disapproval. (9.1.4)
(After the BRM the Convenor dutifully went back and “corrected” his earlier blog post to reflect how the BRM actually operated.)
The Secretaries General further dismiss the concerns regarding BRM voting procedure, saying:
4e. Not correct. Decisions on the comments not discussed during the BRM and proposed dispositions were taken by a process agreed by the BRM itself (29 votes in favour, none against and 2 abstentions).
On the contrary, the BRM was not competent and had not the mandate to set its own voting rules or to negate the provisions for consensus stated in JTC1 Directives 1.2:
These Directives are inspired by the principle that the objective in the development of International Standards should be the achievement of consensus between those concerned rather than a decision based on counting votes.
[Note: Consensus is defined as general agreement, characterised by the absence of sustained opposition to substantial issues by any important part of the concerned interests and by a process that involves seeking to take into account the views of all parties concerned and to reconcile any conflicting arguments. Consensus need not imply unanimity. (ISO/IEC Guide 2:1996)]
The Directives specify the rules. If NB’s do not like the rules, then NB’s may work with SWG-Directives to define new rules and then vote on them using the defined process. But if the rules are not applied correctly, then the proper course is for NB’s to appeal against the actions or inactions of those with a duty to carry out the rules. This is the essential governance model of JTC1. NB’s rule, but they rule through the rules. We may not merely decide by majority vote to ignore rules for this DIS, or to institute new rules for that DIS, or to substitute different rules for another DIS, in an ad-hoc fashion, based on a BRM vote.
Using the logic given by the Secretaries General, what in principle would prevent a BRM from voting itself an Augur in addition to a Convenor for the purpose of observing the flights of birds to decide whether a given change to the DIS text was auspicious or not? Is there any voting procedure that would not be permitted them once we say that a BRM, by majority vote, can institute their own voting rules? Are TMB/SMB certain that this is the principle that they want affirmed by their rejection of the NB’s appeals?
Further, NB’s were not duly notified that their BRM delegations would be determining their own voting rules, so few if any of them had NB instructions on that matter. An agreement among BRM HoD’s to set aside cardinal principles of JTC1, in the absence of NB consultations, should not be allowed to stand.
Finally, the existence of a vote at the BRM is not incompatible with the assertion that the BRM was “too short, arbitrarily short, or otherwise incorrectly conducted”. When given the choice between several bad alternatives, the delegations made a choice. That does not legitimatize the flawed application of JTC1 process that incorrectly gave them only bad choices and forced upon them a vote which they did not have the mandate to hold.
I could go on and on, but I’ll spare you all more of the same. I am sorry to report that I find the response by the Secretaries General to be perfunctory, poorly reasoned and self-serving. It does not serve to resolve the issues, including important issues where clarification is needed. Majority rule, within the rules, should be encouraged. But to dismiss legitimate complaints by arguing that the majority agreed to not follow the rules, this is to substitute mob rule (or orchestrated monopoly rule) for the rule of law. We know where that leads to — curtailed rights for those with minority opinions. And that should concern everyone.
The Secretary General of ISO, Alan Bryden, retires at the end of the year. August vacation is approaching, and before you know it there will be a retirement party with the cake and gifts, maybe a wall plaque or pewter paperweight. I am sure he does not need or desire to spend more time being reminded of the OOXML disaster that occurred during his last year at ISO. TMB/SMB members all want vacation as well. So do I. But out of respect for Mr. Bryden’s eventual successor, and our shared mission in JTC1, shouldn’t we urge TMB/SMB to do their job and not leave this all unresolved for the next guy to deal with? Dismissing an appeal with so many open unresolved issue is not expediency. It is merely creating more dissent, more distrust and more trouble that we’ll all need to deal with next time around. It is better, I think, to hear the appeals, get to the bottom of this, seek resolution, consensus and closure, and then to move on. Ignoring mistakes will not make them go away.
Watch closely to see whether Mr Bryden is offered a well-paying job at CompTia or some other MS Lobbying firm.
Perhaps Mr Bryden will take a senior management position at MS itself in their newly formed ‘Standards Office’ due to his ‘familiarity’ with ISO processes.
There weren’t any mistakes made during the process. Clearly it was the intent of those in authority to force acceptance of OOXML no matter what. They deliberately manipulated procedures to ensure acceptance of OOXML and they succeeded, they made no mistakes achieving their goals.
Continuing to recognize ISO as a standards body, that is a mistake.
kawabago, I disagree.
ISO is still a standards body, and in general has a fairly decent record of setting good standards.
Which is, itself, even more reason why these appeals should be thoroughly investigated, and not dismissed out-of-hand as the Secretariat would like.
I think kawabago is right.
Yes ISO has a good track record for setting standards. But this fiasco shows there is no accountability in ISO. Once it is subverted, there is no mechanism to straighten it up. Although past standards might be OK, we can’t be confident about future standards. Microsoft has shown the way to subvert ISO and the road is wide open to copycats.
Can somebody please explain how this vote actually works? How should a voter proceed if, say, he thinks that two of the three appeals should be processed further? (Let’s assume he supports the ABNT and BIS appeals, but not FONDONORMA or SABS. I just picked these based on alphabetical order.)
Remember, the instruction says to vote on either a) OR b) (that’s as close as English comes to specifying an XOR, so I believe the intent is that you can’t vote on both a) and b) for the same appeal).
I guess you could vote on a) for each of the appeals: a NO vote for each of ABNT and BIS (to signify you DO want the appeal to proceed) and a YES vote for each of FONDORAMA and SABS.
Alternatively, you could vote on b); a YES on each of the first two and a NO on each of the second.
You might try voting on both a) and b) but that might get your ballot invalidated.
Now, how are the ballots going to be counted? Say that the majority of voters who vote on ABNT in question a) vote YES, and that the majority of voters who use question b) to vote on ABNT also vote YES. Since a) and b) give opposite results, they logically can’t both pass on the same ballot, but this strange voting structure makes this result possible.
I noticed that as well. They managed to turn 4 questions into 8, while reversing the sense of the question they should be asking.
What really should be asked is:
1. Should the appeal of ABNT be processed further (Yes or No)?
2. Should the appeal of BIS be processed further (Yes or No)?
3. Should the appeal of FONDONORMA be processed further (Yes or No)?
4. Should the appeal of SABS be processed further (Yes or No)?
That approach is so natural and common that I do wonder what they are up to with the twisted way the ballot was created. Essentially they’ve introduced additional degrees of freedom into the ballot that don’t map into distinctions relevant to the questions that the Directives asked them to decide. It could just be ineptitude. But this is coming from the Secretaries General of ISO and IEC. I’d expect they know how to create a ballot.
“Degrees of freedom” is a good choice of words. It’s an old tactic. With a confusing enough ballot, there are good odds of a self-contradictory result, or at least a result that could be interpreted multiple ways. That leaves the secretariat free to act how they like.
I’m wondering how important ISO is for these types of ICT standards anyway? Almost all of the internet happened despite anything they did (tcp/ip, http, html, smtp, mime, jpeg, png, dns, s/mime, etc). Many other commercially important ICT (or related) standards are consortia/vendor based or de-facto (posix, cd, dvd, audio, video, usb, firewire, cardbus).
In the mean-time ISO gave us the OSI network stack (i think?), and SGML (how important commercially is it really?). Real winners there?
Unicode (in particular) and C perhaps, but how much of that was ISO and how much was just rubber stamped by them?
I certainly remember during the mid 90’s how ISO was almost completely irrelevant to the creation and growth of the internet and computer systems in general. Why are they suddenly important now?
I’m just starting to wonder if all this stress is worth it. Although having that ISO rubber-stamp might mean it gets considered, any organisation (or govt) will have to deal with it being a useless standard if they are foolish enough to make decisions based soley upon that stamp.
Rick Jelliffe says
notzed: ISO SGML is a very successful ISO standard: you perhaps use it under the name (or brand) “XML”. Incredibly important.
Many of the things you mention are in fact ISO standards already: JPEG, PNG, the common video standards, POSIX ABI (and Linux!).
You mentioned ISO C and Unicode (actually, Unicode Character Set comes from a consortium which adds extra information to the IS10646 character list, but it is frequently the same I18n people in IS10646, Unicode Consortium, W3C I18n wearing different hats.)
Firewire is not a de facto standard, but an IEEE standard (IEEE is the international standards body that looks after hardware interfaces).
But your premise seems to be that a standard is important if it gets mass adoption and never changes. However, some standards are intended for niche uses and for particular industrial sectors only. And technologies are not static but evolve and cross-pollenate: for example, LDAP grew out of X.500 which grew out of X.400.
As far as people being foolish enough to make a decision based on a stamp, this is in fact exactly what ISO (and the other bodies) themselves say: these are voluntary standards and the responsibility for evaluating whether they are appropriate to use in any situation is the adopters’. You seem to think you are saying something shocking…
All the stamp does is say that the documents have been through a formal development or QA process and successfully been voted for by the interested National Bodies: but standardization takes the form of a check of aspects of documentation and is not any indication that the technology described is necessarily preferred, optimal or even adequate for any specific task.
The ISO standard for Torx screws, for example, does not mean that you cannot use Phillips head screws. But if you do want to use them, as mature ‘stabilized’ standards, they have an adequate description (and if the description is not adequate, they have a formalized process for getting the description improved.)
It is a fair point. The XML work done by JTC1 is minuscule in comparison to the number and importance of XML standards done by other standards organization with less pedigree. For example, the W3C, OASIS, or the large number of industry-oriented consortia.
A fair question might be why this is so? There is nothing to prevent a company or individual from petitioning their NB to create a new work proposal in SC34 for the next great XML standard. But this rarely happens.
Even when this rarely happens apperantly it it much more likely to happen when the EU asked organization to do so.
So we’re “down to the last week” means that there was a decision to make, yes? Is the result of the decision public?
I believe August 4th is the final date for TMB/SMB to vote on whether to hear the appeals further.
There is nothing in the process that requires the results to be made public. However, since ISO’s press office decided to publicize the lodging of the appeals, I assume they will also issue a press release if/when the appeals are resolved.
From what I hear (indirectly) the appeals are getting a lot of support.
It is well past the 4th, anyone know what happened?
Was the vote to proceed and that is why ISO is so quiet?
I’m hearing conflicting reports, including one that says the vote was tied, but nothing official yet. I’d expect an officious ISO press release within a few days.
ISO has just published this:
“The two ISO and IEC technical boards have given the go-ahead to publish ISO/IEC DIS 29500, Information technology – Office Open XML formats, as an ISO/IEC International Standard after appeals by four national standards bodies against the approval of the document failed to garner sufficient support.”
Luc Bollen says
A press release has now been published by ISO and IEC, and as expected it is a farce.
See also http://www.noooxml.org/forum/t-81761/not-very-appealing:iso-iec-kicks-out-appeals
You guys lost and you’re still whining about it, and today I read that you (IBM) are going to form your own ISO-rival? LOLOLOLOLOL
Wow, talk about a bunch of cry babies!!
Microsoft joined the OASIS ODF tech committee, which you CLAIMED you wanted to happen (we all know that was bullshit, you were only saying that for show). Once they start taking part in ODF discussions, I can’t wait for them to shove their spec needs down your throat and watch you try to reject them, which will put the lie to your claims of ODF as one spec to rule the rest.